This TRD, Hardware and High Field, has been in existence since 2010 and has been addressing the demands for increased spatial resolution, sensitivity and speed as well as solving problems of high magnetic field, by developing novel MR hardware. We propose to continue this focus on technology development, particularly focusing our work on solving neuroimaging problems at high field (3T) and ultra-high-field (7T and above). With regard to the latter, it is known that there are still substantial technical innovations needed to make ultrahigh- field MRI routine, stable and consistently superior to the best available clinical MRI systems, in all parts of the body. The technical challenges related to gradient, shim and RF performance, decreased B0 and B1 homogeneity, and increased RF power deposition are the most critical. These challenges are the basis for much of the present research activity in the UHF MRI world, and many creative solutions are being found. But one underlying principle is clear: solving these problems will demand innovation in the design, implementation and application of high-performance hardware sub-subsystems. It is also clear that even at field strengths lower than 7T, many improvements in image quality would be enabled through novel hardware development. From the Human Connectome Project comes a clear demand for increased gradient performance, which is needed both for more efficient diffusion encoding and for faster and higher resolution spatial encoding. Yet body-size gradients have now reached hard amplitude and slew rate limits set by human peripheral nerve stimulation thresholds, and therefore any further increases in gradient performance will require innovation in smaller size gradient coils, most obviously head-size gradients. Along with the demands for better gradients come new requirements for better B0 shimming and B1 / radio frequency performance. In this TRD project, we will pursue projects involving major hardware design, construction and analysis in all three of these principal hardware subsystems of the MR scanner.